Effects of fourth ventricle bombesin injection on meal-related parameters and grooming behavior

Effects of fourth ventricle bombesin injection on meal-related parameters and grooming behavior

Peptides, Vol. 12, pp. 761-765. ©Pergamon Press plc, 1991. Printed in the U.S.A. 0196-9781/91 $3.00 + .00 Effects of Fourth Ventricle Bombesin Injec...

527KB Sizes 0 Downloads 17 Views

Peptides, Vol. 12, pp. 761-765. ©Pergamon Press plc, 1991. Printed in the U.S.A.

0196-9781/91 $3.00 + .00

Effects of Fourth Ventricle Bombesin Injection on Meal-Related Parameters and Grooming Behavior F R A N C I S W. F L Y N N

Department of Psychology and Neuroscience Program, Box 3415 University Station University of Wyoming, Laramie, WY 82071 R e c e i v e d 25 January 1991 FLYNN, F. W. Effects of fourth ventricle bombesin injection on meal-relatedparameters and grooming behavior. PEPTIDES 12(4) 761-765, 1991.--Injections of bombesin (BN) into the vicinity of the caudal brainstem suppress food intake in rats. In the present study, the food intake parameters [meal size (MS). intermeal interval (IMI), satiety ratio (SR)] affected by 4th ventricle BN injections were determined. Following a 15-h food deprivation, rats were administered 4th ventricle injections of saline (0.15 M) and BN in doses of 1, 5, 10, and 20 ng BN, and were then given access to sweetened milk. The animals' behaviors (feeding, resting, grooming, exploring) were scored every one min and milk intake every five min for 60 min following the injections. Fourth ventricle injections of 5 ng BN and greater reliably suppressed milk intake. This reduction was reflected in a significant reduction in the MS. The IMI was not affected. As a result, the SR (IMI2/MS~), which is thought to represent the satiating property of food, was reliably greater following BN than following saline administration. The reduced food intake was accompariled by a significant increase in grooming behavior and a corresponding decrease in exploring. The amount of time spent resting (inactive) was similar following saline and all but the highest dose of BN. To demonstrate that the behavioral effects of BN were mediated by specific caudal brainstem BN receptors, 4th ventricle injections of [D-Phel2,Leut4]BN, a BN receptor antagonist, or saline preceded the 4th ventricle injections of 5 ng BN. Pretreatment with [D-Phe~2,Leu~4]BN reliably blocked the effects of BN on food intake and grooming. Bombesin

Food intake

Brainstem receptors

Grooming behavior

food intake can provide information as to the underlying mechanism of BN action. For example, the reduction in food intake following systemic injections of BN reflects both a decrease in MS and an increase in the IMI (21). The effect of systemic BN on reducing MS is abolished following combined dorsal rhizotomy and vagotomy but the increase in the IMI is not affected by this visceral neural disconnection of the gut from the brain (21). Thus the effects of BN on MS and IMI are mediated by separate neural systems and the effects of BN on IMI may reflect a central site of action. Since the caudal brainstem is highly sensitive to BN and 4th ventricle BN injections suppress food intake, it was of interest to determine the meal parameters that are modified by 4th ventricular BN administration. Furthermore, to evaluate the hypothesis that the behavioral effects of 4th ventricle BN administration are mediated by specific BN-like peptide receptors and do not reflect nonspecific action, rats were pretreated with 4th ventricle injections of [D-Phe12,Leu14]BN, a specific BN receptor antagonist (9,15), and food intake and behavior were monitored.

BOMBESIN-LIKE (BN) peptides include bombesin, which was originally isolated from amphibians, and several mammalian peptides that have a strong C-terminal homology to BN (4, 14, 16). These mammalian BN-like peptides include gastrin releasing peptide (GRP), which shares the greatest sequence homology to amphibian BN (14). BN-like peptides influence a number of biological systems (3,6), including food intake. For example, systemic injections of BN and GRP inhibit food intake under a number of conditions in rats, but do not inhibit water intake by water-deprived rats (7, 8, 19). The observations that BN and GRP injections suppress food intake in decerebrate rats (6) and that 4th ventricle injections of BN-like peptides also suppress feeding [(5, 12, 13); Flynn, unpublished observations] raise the possibility that BN-like peptides may act directly on caudal brainstem systems to control ingestive behavior. Also, 4th ventricle BN injections have no direct effect on water intake by water-deprived rats (5,12) and rats do not acquire an aversion to tastes paired with 4th ventricle BN injections (5). Reports that direct application of BN into the nucleus of the solitary tract suppresses food intake (2,10) are consistent with the suggested role of caudal brainstem BN-like peptides in the control of food intake. In previous studies, only cumulative 15-, 30-, and 60-min intake following 4th ventricle BN injections was measured (5,12). The amount of food consumed is a function of the meal size (MS) and intermeai interval (IMI). The effects of BN on feeding can therefore be characterized by examining the microstructure of feeding. Furthermore, analysis of the microstructure of

METHOD

Subjects and Surgery Male Sprague-Dawley rats ( n = 9 ; approx. 300 grams) were housed in standard wire-mesh cages in a temperature-controlled colony room with a 12-h light-dark cycle. The animals had access to Purina pellet chow and tap water unless otherwise indicated.

761

762

Rats were anesthetized with Nembutal (50 mg/kg, IP), mounted in a stereotaxic instrument, and their skulls exposed. The guide cannula (26 gauge; Plastic One Co.) entered the brain through a hole drilled in the skull on the midline, 2.5 mm anterior to the interparietal-occipital suture. The dura was cut and the guide cannula was lowered 5.5 mm ventral to the dura. The carmula was then cemented to four screws anchored to the skull and sealed with an obturator. These coordinates position the guide cannula just dorsal to the roof of the 4th ventricle (5). Following surgery, rats were treated with 10,000 units of bicillin IM and allowed to recover for one week with ad lib access to Purina chow and water.

Procedure Rats were adapted in stages to the test procedure. The rats were first familiarized with the clear plastic test chambers (45 × 24× 19 cm) by being placed in the chambers and left undisturbed for 0.5 h. Next, food was removed prior to the onset of the dark portion of the light/dark cycle. Following a 15-h deprivation, each rat was removed from its home cage, the obturator was removed, and an injector was inserted into the cannula. The injector extended 1 mm beyond the tip of the cannula and into the 4th ventricle. Rats were administered 5 ~,1 sterile isotonic saline at a steady rate over 1 min using a hand-held microliter syringe. The obturator was then replaced and the rats were then placed in the test chambers. Calibrated bottles were filled with sweetened milk (equal parts of sweetened condensed milk and water) and then attached to the fronts of the chambers. Intake was measured (nearest ml) 1 h later. Rats were adapted to this procedure three times prior to the start of testing. The same procedure was used during testing with the exception that, following the 4th ventricle injection, direct behavioral observations were made once a minute for 60 min for each rat. Rats were administered, in a random order, 4th ventricle injections (5 Ixl) of 0.15 M saline or BN [1, 5, 10, 20 ng (Sigma Chemical Co.)]. Each minute following the injection the observer noted the behavior being displayed by the rat and recorded the behavior as either resting (inactivity), exploring (locomoting, rearing), grooming, or eating. The amount of diet consumed was recorded every 5 min for the 1-h period. The MS was defined as the amount of food consumed before the occurrence of 5 consecutive observations of noneating behavior. From the behavioral assessment the IMI was defined as the time between the termination of the first meal and the initiation of the subsequent observation of feeding (20). If there was no second meal, the IMI was defined as the time between the termination of the fast meal and the end of the test period (20). Following the 1-h test, rats were returned to their home cages with ad lib access to Purina chow and water. A minimum of two days separated successive drug administrations. Bombesin antagonists. Next, rats were pretreated with 4th ventricle injections of the BN receptor antagonist [D-Phe12,Leulg]BN (9,15). The same test procedure as described above was used. Rats were administered a 4th ventricle injection of 500 ng [D-pbeI2,Leu14]BN and approximately 3 min later a second injection of either saline or 5 ng BN, a dose that was found to reliably reduce food intake. All injections were a 5 ~xl volume. Rats were then placed in the test chambers and behaviors and intake monitored for 1 h as before.

Histology Rats were deeply anesthetized and India ink (5 Ixl) was injected through the 4th ventricle cannulas. The animal was then perfused successively with isotonic saline and 10% buffered formalin. The brain was extracted, cut in the midsagittal plane, and

FLYNN

examined using a dissecting microscope. The position of the cannula tract and the area in which ink was deposited were recorded. RESULTS

Histology Ink was deposited into the 4th ventricle of 9 rats and it did not spread rostral to the recess of the inferior colliculus. The cannula tract was just anterior to the primary cerebellar fissure and both the placement of the cannula and the extent of ink spread were similar to that previously reported (5). In these 9 animals there was no damage to the dorsal medulla. For each of these 9 animals, the MS, IMI, SR [IMIz/MS ~ (17)], and the amount consumed (in ml) were computed. The SR was calculated because it provides information on the effectiveness of food to inhibit subsequent feeding or the satiating capacity of the food (17). The times spent grooming, resting, and exploring were also tabulated. Separate, repeated measures analyses of variance were used to examine the effects of the 4th ventricle injections on each of the behavioral parameters. Further comparisons were made using Newman-Keuls tests.

Food Intake In general, rats began eating within a short time of being placed into the chamber. More specifically, rats initiated eating on average 1 min [+-0.3 min (-S.E.M.)] following 4th ventricle injections of saline, 1, 5, 10 ng BN, and 5 ng BN + [D-Phe~2,Leu~4]BN. In contrast, the latency to initiate eating was longer following 20 ng BN (21-4-8 min) than following saline (p<0.05). As indicated in Fig. 1, food intake at 15, 30, 45, and 60 min following injections of 5, 10, and 20 ng BN was rehably less than that following saline, F(5,40)= 10.4, p<0.0001 (Fig. 1). These doses of BN were equally effective in suppressing food intake. The lowest dose of BN, 1 rig, did not suppress food intake. Injections of [D-Phel2,LeuV*]BN blocked the effects of 5 ng BN on food intake and the amount of food consumed was not reliably different from that following saline injections.

Meal Size and lntermeal Interval The size of the first meal was computed for each of the injection conditions. Compared to 4th ventricle saline injections, injections of 1 ng BN had no reliable effect on MS. In contrast, the size of the first meal was reliably reduced by 4th ventricle injections of 5 ng BN and greater, F(5,40)=6.9, p<0.001. Following these injections MS was reduced on average 50% (Fig. 2). The effects of 5 ng BN on MS were blocked by the 4th ventricle injections of [D-Phel2,Leula]BN. Furthermore, the MS following the combined treatment of the antagonist and 5 ng BN was reliably greater than that recorded following 5, 10, and 20 ng BN (p's<0.05), and not reliably different from the MS following saline injection. The IMI following 4th ventricle BN (25 +- 3 min) injections was not reliably different from that following saline injections (18___3.7 min).

Satiety Ratio The SR following saline injections was similar to normal daytime ratios (17). Central injections of BN reliably increased the SR (IMI2/MS~) compared to saline, F(5,40)=3.1, p<0,01 (Fig. 3). Furthermore, only those injections that reliably reduced food intake (5, 10, and 20 ng) increased the SR (p's<0.05). Fourth ventricle injections of [D-Phe12,Leu14]BN reliably blocked the effect of 5 ng BN on the SR. Also, the SR following the combined treatment of the antagonist and BN was not

MEAL-RELATED PARAMETERS AND BOMBESIN

20

0 • • • • ql.

18

18

763

.qallr, a

Saline I ng BN 5 n g BN 10 n g BN 20 n g BN D - P h e + 5 n g BN

14

A

z 6 t~ °,..~

4

2

I

0 0

,

15

TIME

,

I 30



I

I

I

4-5

I

I 60

(Minutes)

FIG. 1. Cumulative intake (mean---S.E.M.) following 4th ventricle injections of saline, 1, 5, 10, 20 ng BN and pretreatment with [D-PheI2,LeuI4]BN prior to administration of 5 ng BN. (Food intake following central injections of 5, 10, and 20 ng BN was significantly less than following saline, and the suppressive effect of 5 ng BN on food intake was blocked by pretreating the animal with 4th ventricle [D-Phet2,Leul4]BN.)

FIG. 2. The amount consumed during the first meal was significantly reduced following 4th ventricle injections of 5 ng BN and greater and the suppressive effect of 5 ng BN was blocked by the 4th ventricle injection of [D-PheI2,LeuI4]BN. [Asterisk indicates significantly different from saline (p<0.05).]

reliably different from that following saline injection.

Grooming, Exploring and Resting Fourth ventricle injections of 5, 10, and 20 ng BN reliably increased the duration of grooming compared to saline, F(5,40)=27, p<0.001 (Fig. 4). The duration of grooming following saline and 1 ng BN was not reliably different. Also, injection of [D-Phe12,Leu14]BN completely blocked the stimulatory effect of 5 ng BN on grooming. Rats displayed reliably more exploring (locomoting, rearing) following saline and 1 ng BN than following 4th ventricle injections of 5, 10, and 20 ng BN, F(5,40)= 7.4, p < 0 . 0 5 (Fig. 4). The amount of exploration that was displayed by rats pretreated with [D-PheI2,Leu14]BN was not reliably different from that observed following saline, but was reliably greater than that recorded following injections of BN (p's<0.05). As such, the antagonist completely blocked the effect of BN on exploration. As shown in Fig. 4 (lower panel), 4th ventricle injections of 20 ng BN reliably reduced the time that the animal spent resting (inactive) compared to saline (p<0.05). Except for the decrease observed following 20 ng BN, animals were inactive for similar amounts of time following saline, 1, 5, 10 ng BN, and when pretreated with [D-Phe12,Leul'*]BN.

6 m

m

Saline 1 n g BN

~I~

5 n g BN 10 n g BN 2 0 n g BN D - P h e + 5 ng BN

v

o

__L ¢0

Sequence of Behaviors The temporal relationship of eating, grooming, resting, and exploring is shown in Fig. 5 for saline, the midrange dose of BN (10 ng) and 20 ng BN, which, unlike the lower doses, was

FIG. 3. Fourth ventricle injections of BN reliably increased the SR and this increase in the SR was blocked by [D-Phe12,Leu14]BN. [Asterisk indicates significantly different from saline (p<0.05).]

764

FLYNN

ioo

60

E o

40

Saline i ng BN 5 ng BN

80

[~

I0 ng BN 20 ng BN

6O

1

D-Phe + 5 ng BN

0

~

40

~ o .r-i

2o

2o

SALINE

o Exploring

\

¢~ ~oo liQ

10 n g BN

80

og~ 0

@1"3Eating Grooming 0 Renting

.~

~

SO

/o\ o

0

_~ 4o

3

2o

o IO(I

"~ ._.

2o

20 n g BN

0

4O 0

20 5

10

15

20

25

30

35

40

45

50

55

60

TIME (min)

co, (;0 f 40 o

FIG. 5. The temporal occurrences of eating, resting, grooming, and exploring behaviors are expressed as a percentage of total observed behaviors during 5-rain intervals after 4th ventricle injections of saline (upper panel), 10 ng (middle panel), and 20 ng BN (lower panel).

iv,

= 2o °-l-

FIG. 4. The panels display the time that the animals engaged in: grooming behavior (upper panel), exploring (locomoting, rearing; middle panel), and resting (inactivity; lower panel). Following injections of 5 ng BN and greater, exploring is replaced by grooming behavior but the time spent resting is not significantly affected by 4th ventricle BN injections of 1, 5, and l0 ng.

found to decrease resting behavior. The frequency of each behavior during 5-rain intervals is expressed as a percentage of total observed behaviors (20). Immediately following 4th ventricle saline injections, eating and exploring dominate the behaviors and grooming and resting behaviors are largely absent. As time progresses, eating decreases and grooming and resting increase. By 60 min, resting behavior is predominant over all others. As observed following saline, eating is a predominant behavior immediately following 4th ventricle injections of 10 ng BN. Thereafter, eating decreases and resting behavior increases. Resting is the predominant behavior from 45 min until the end of the observation period. Unlike saline, however, grooming behavior is present from the onset of the observation period and persists throughout the 1 h, with a corresponding decrease in exploring during the same time period. Following injections of 20 ng BN, grooming occupies even a greater percent of total observations and is the predominant behavior at every time point. DISCUSSION In general, 4th ventricle injections of BN reliably suppressed food intake and did not delay the initiation of eating. This is consistent with the previously reported effects of 4th ventricular

BN injections (5,12). Only the highest dose, 20 ng, reliably delayed the initiation of feeding. The change in food intake was shown to reflect a specific change in the microstmcture of feeding. More specifically, the reduction in food intake following 4th ventricle BN injections was accomplished by a decrease in the MS. An increase in the IMI was not required for the reduction in food intake since 4th ventricle BN injections had no reliable effect on the IMI. Similarly, intralateral hypothalamic BN injections reduce MS (20). Therefore, the reduction in MS is a feature common to centrally administered BN. The food intake parameters modified by 4th ventricular BN injections, however, do not encompass the parameters that are affected by systemic BN injections. As noted earlier, systemic injections of BN reduce the MS as well as increase the IMI (21). As well as affecting the MS, 4th ventricle BN injections increased the SR. The increase in the SR suggests that the food consumed during the first meal was more satiating following central injections of BN than following saline. Behavioral analyses indicate that, at least in food-deprived rats, 4th ventricle BN injections increase grooming behavior with thresholds similar to those required to suppress feeding. This raises the possibility that changes in food intake are secondary to increases in grooming. There is some precedent for this position; lateral ventricular BN injections elicit very large increases in grooming behavior and corresponding reductions in resting behavior and food intake (8,11). While increases in grooming appear to be responsible for the reductions in food intake following lateral ventricle BN injections, the same is not necessarily true for 4th ventricle BN injections, especially lower doses of BN. For example, unlike lateral ventricle BN injections, 4th ventricle BN injections in concentrations up to and including 10 ng do not interfere with resting; rather, grooming behavior replaces exploratory behaviors (e.g., locomoting, rear-

MEAL-RELATED PARAMETERS AND BOMBESIN

765

ing). Also, the increase in grooming following 4th ventricle BN injections does not affect the IMI. Finally, the behavioral sequences following injections of saline and 10 ng BN share several similarities. In particular, rats initially engaged in eating and activity (exploring or grooming) and as time progressed, resting became the predominant behavior. This sequence is considered to be indicative of satiety (1). A different behavioral sequence was present following the highest dose of BN; grooming behavior predominated throughout the 1 h. The reduced food intake following 20 ng BN would thus appear to reflect the predominance of grooming behavior. A previous report indicated that 4th ventricle BN administration suppressed food intake at doses lower than those required to elicit atypical activity patterns (12). Higher doses of BN were accompanied by both decreased feeding and increased grooming (12). One interpretation of these results and those of the present study is that while food intake is inhibited by 4th ventricle BN injections, the underlying mechanism (satiety, competing behaviors) responsible for the inhibition of food intake is dose dependent. That is, lower doses may mimic internal signals related to satiety but higher doses recruit competing responses that are incompatable with eating, which contribute to the reduction in food intake. The [D-Phe12]BN analogues function as specific receptor antagonists (9). Lateral ventricular injections of an equally potent [D-Phe~2]BN receptor antagonist partially blocked the effects of

centrally administered BN on feeding and grooming (15). In the present study, 4th ventricle application of [D-Phe12,LeuI'~]BN completely reversed the effects of 4th ventricle BN injection on feeding, grooming, and exploring. The ability of the antagonist to completely block the effects of BN is consistent with the notion that intraventricularly administered BN exerts its intakecontrolling effects through caudal brainstem BN-like peptide receptors. Several observations raise the possibility that the nucleus of the solitary tract (NST) is a caudal brainstem site that may mediate the intake-controlling effects of BN administered into the 4th ventricle. The NST contains a high concentration of BN-like peptide binding sites, immunoreactivity, and mRNA (18, 22, 23). Also, local injections of BN into the NST mimic the effects of 4th ventricle BN administration on food intake [(2,10); Flynn, in preparation]. Lastly, damage to the area postrema and underlying NST abolishes the suppressive effects of 4th ventricle BN administration on food intake (13). ACKNOWLEDGEMENTS This research was supported by a grant from the National Institutes of Health (RO1-NS24879) awarded to the author. A preliminary report of these data was presented at the Winter Neuropeptide Conference, Breckemidge, CO, 1991. The valuable assistance provided by Ramiro Ramos and Carla Davis was greatly appreciated.

REFERENCES 1. Antin, J.; Gibbs, J.; Holt, J.; Young, R. C.; Smith, G. P. Cholecystokinin elicits the complete behavioral sequence of satiety in rats. J. Comp. Physiol. Psychol. 89:784-790; 1975. 2. de Beaurepaire, R.; Suaudeau, C. Anorectic effect of calcitonin, neurotensin and bombesin infused in the area of the rostral part of the nucleus of the tractus solitarius in the rat. Peptides 9:729-733; 1988. 3. Brown, M.; Vale, W. Bombesin--a putative mammalian neurogastrointestinal peptide. Trends Neurosci. 10:95-97; 1979. 4. Erspamer, V.; Melchiorri, P. Amphibian skin polypeptides active on the gut. J. Endocrinol. 70:12p-13p; 1976. 5. Flynn, F. W. Fourth ventricle bombesin injection suppresses ingestive behaviors in rats. Am. J. Physiol. 256:R590-R596; 1989. 6. Flynn, F. W. Bombesin-like peptide effects on ingestive behavior and plasma glucose in decerebrate rats. Soc. Neurosci. Abstr. 16: 775; 1990. 7. Gibbs, J.; Fanser, D. J.; Rowe, E. A.; Rolls, B. J.; Maddison, S. Bombesin suppresses feeding in rats. Nature 282:208-210; 1979. 8. Gibbs, J.; Kulkosky, P. J.; Smith, G. P. Effects of peripheral and central bombesin on feeding behavior of rats. Peptides 2(Suppl. 2): 179-183; 1981. 9. Heinz-Erian, P.; Coy, D. H.; Tamura, M.; Jones, S. W.; Gardner, J. D.; Jensen, R. T. [D-Phe ~2] bombesin analogues: A new class of bombesin receptor antagonists. Am. J. Physiol. 252:G439-G442; 1987. 10. Johnson, S. A.; Merali, Z. Specific neuroanatomical and neurochemical correlates of grooming and satiety effects of bombesin. Peptides 9(Suppl. 1):233-244; 1988. 11. Kulkosky, P. J.; Gibbs, J.; Smith, G. P. Behavioral effects of bombesin administration in rats. Physiol. Behav. 28:505-512; 1982. 12. Ladenheim, E. E.; Ritter, R. C. Low dose fourth ventricular bombesin selectively suppresses food intake. Am. J. Physiol. 255:R988R992; 1988.

13. Ladenheim, E. E.; Ritter, R. C. Caudal hindbrain participation in suppression of feeding by central and peripheral bombesin. Soc. Neurosci. Absn-. 15:964; 1989. 14. McDonald, T.; Jornvall, H.; Nilsson, G.; Vagne, M.; Ghatei, M.; Bloom, S. R.; Mutt, V. Characterization of gastrin releasing peptide from porcine non-antral gastric tissue. Biochem. Biophys. Res. Commun. 90:227-233; 1979. 15. Merali, Z.; Merchant, C. A.; Crawley, J. N.; Coy, D. H.; HeinzErian, P.; Jensen, R. T.; Moody, T. W. (D-Phe 12) bombesin and substance P analogues function as central bombesin receptor antagonists. Synapse 2:282-287; 1988. 16. Moody, T. W.; Thoa, N. B.; O'Donohue, T. L.; Pert, C. B. Bombesin-like peptides: Localization in synaptosomes and release from hypothalamic tissue. Life Sci. 26:1707-1712; 1980. 17. Panksepp, J. Reanalysis of feeding patterns in the rat. J. Comp. Physiol. Psychol. 82:78-94; 1973. 18. Panula, P.; Yang, H.; Costa, E. Neuronal location of the bombesinlike immunnreactivity in the central nervous system of the rat. Regul. Pept. 4:275-283; 1982. 19. Stein, L.; Woods, S. Gastrin releasing peptide reduces meal size in rats. Peptides 3:833-835; 1982. 20. Stuckey, J. A.; Gibbs, J. Lateral hypothalamic injection of bombesin decreases food intake in rats. Brain Res. Bull. 8:617--621; 1982. 21. Stuckey, J. A.; Gibbs, J.; Smith, G. P. Neural disconnection of gut from brain blocks bombesin-induced satiety. Peptides 6:1249-1252; 1985. 22. Wada, E.; Way, J.; Lebacq-Verheyden, A.; Battey, J. F. Neuromedin B and gastrin releasing peptide mRNAs are differentially distributed in the rat nervous system. J. Neurosci. 10:2917-2930; 1990. 23. Zarbin, M.; Kuhar, M.; O'Donnhue, T.; Wolf, S.; Moody, T. W. Autoradiographic localization of (125I-TYR4) bombesin-binding sites in the rat brain. J. Neurosci. 5:429-437; 1985.